Skip to content

Pandemic Puzzle: The role of data in the pandemic

Stanford Medicine biomedical data scientist weighs in on the role of data as we respond to the pandemic and prepare for the future.

On Oct. 28, epidemiologists, public health advocates, government leaders and technology experts gathered to discuss how to improve our national response to the COVID-19 pandemic.

The virtual symposium -- "Tracking and Mitigating a 21st Century Pandemic" -- was the third in a four-part Pandemic Puzzle series, hosted by the Stanford School of Medicine and the  Stanford Graduate School of Business, that examines how the United States -- and the world -- responded to the health crisis.

In the day's first panel, participants discussed the many challenges faced by the United States public health system as it strained to monitor the emerging pandemic and identify appropriate interventions to prevent disease spread. I caught up with panelist Nigam Shah, MBBS, PhD, professor of medicine and of biomedical data science at Stanford Medicine, after the session to talk about the unique role of data and data scientists in counteracting infectious diseases.

What stumbling blocks did data scientists at Stanford Medicine encounter early in the pandemic? Are these issues still a problem today?

In the early days, the number of patients hospitalized with COVID-19 on any given day would differ depending on what source you consulted. We jokingly referred to the situation as the "Disunited States of Data." Most of the issues could be traced to differing definitions or delays in data movement (of test results or hospital admission status, etc.).

Additionally, each data-use agreement had to be negotiated from scratch and ad hoc contract negotiations were needed to allow our faculty access to the county's data systems and establish the legal and computational framework for transferring data securely. It often took weeks to negotiate contracts when both parties were willing and eager to partner.

A lot of these issues have been streamlined as a result of joint efforts of the public health officials, the American Hospital Association and state-level efforts. However, other problems -- such as the lack of standardized ways to report public health data or to specify the conditions under which data can be shared broadly for public good, and with whom -- persist.

You mentioned that data needs to be treated as a 'first-class citizen' in future public health efforts. What role can informatics professionals play?

Unfortunately, public health reporting is a secondary and separate activity from the point of view of our electronic data systems, which are designed for tracking patient care and facilitating billing. The need for county-, state- and federal-level public health reporting are not even considered when these systems are built, although it relies on the same data.

We should collect, store and disseminate data once, in a way that is useful to all. It shouldn't be necessary for a human to manually summarize data (say the number of admitted patients with COVID-19) and manually enter it into a smartsheet for public health reporting. Informatics professionals know how the data pipes are connected inside of health systems and can help with these efforts.

How can data acquisition and aggregation be optimized to improve public health?

It's important that the form in which public health data are collected and reported is agreed upon nationally. Think of our electrical grid. Imagine how much trouble it would be if each county had its own shape and size of electric outlets. Having an agreed upon standard is a necessity. This isn't a new idea -- members of the American Medical Informatics Association were advocating for wider adoption of data standards long before I finished graduate school -- but it's been difficult to implement.

We cannot overcome decades of neglect of public health with a few months of meetings and a few bills. It will take sustained effort to build a good infrastructure for public health. But progress is being made. For example, one of our faculty, Mark Musen, MD, PhD, is an international leader in the creation of innovative technology to standardize datasets.

But we need to agree on the collection processes and drive mass adoption so that everybody uses the same standard, analogous to our national agreement that our electrical voltage is 110V, and that the plug fits into the socket with two flat pins and a round grounding pin.

In a perfect world, how would we handle the next pandemic?

The one thing that remains true is that, despite all the stumbles and goof ups, we as a community have learned and will continue to learn. Perhaps not fast enough and not deep enough, but everyone has learned something about how we handle the next one.

What we need is national leadership, perhaps by an empowered Centers for Disease Control or at the level of the Department of Health and Human Services. I'd urge our elected leaders to include in our revised national pandemic preparedness plan guidance on the roles of companies, universities, local and state health officials and pre-agreed upon "rules of engagement."

At the very least we can begin by speaking the truth from day one. For example, don't say, "Masks are not needed." Instead say, "We think masks would help, but we are short on supply and will get them to you in a few days."

The fourth and last installment of "The Pandemic Puzzle: Lessons from COVID-19" on Nov. 19 will bring together leading immunologists, public health experts and government and industry professionals to discuss innovation and discovery during a global crisis. You can register for free here.

Photo by Martin Sanchez

Popular posts

How the tobacco industry began funding courses for doctors

Earlier this year, the largest tobacco company in the world paid millions to fund continuing medical education courses on nicotine addiction —16,000 physicians and other health care providers took them.